<i>k</i>-means clustering of the significant transcript lists reveals similar clusters.

<p>Mean normalised expression profile of individual clusters (y axis, ±SD, n = 4) at each time point (x axis; 0, 1, 3, 6, 12 and 24 hours) for each cluster. N: number of transcripts within cluster; P: most significant canonical pathway (IPA). Clusters are grouped by similarity in kinetic profile (Pearsons correlation, <a href="http://www.plosone.org/article/info:doi/10.1371/journal.pone.0097702#pone.0097702.s007" target="_blank">Table S2</a>) and top canonical pathway.